Home - News ( United States | United Kingdom | Italy | Germany ) - Football scores

Showing content from https://patents.google.com/patent/TWI321316B/en below:

TWI321316B - Compression unit and method for compression of parameters, a decoder and a method for decoding encoded blocks of parameters and a computer program

Publication number

TWI321316B

TWI321316B TW095112773A TW95112773A TWI321316B TW I321316 B TWI321316 B TW I321316B TW 095112773 A TW095112773 A TW 095112773A TW 95112773 A TW95112773 A TW 95112773A TW I321316 B TWI321316 B TW I321316B

Authority

TW

Taiwan

Prior art keywords

tuple

parameter

sequence

parameters

parameter set

Prior art date

2005-04-13

• Espacenet
• Global Dossier
• Discuss

• 230000006835 compression Effects 0.000 title claims description 39
• 238000007906 compression Methods 0.000 title claims description 39
• 238000000034 method Methods 0.000 title claims description 34
• 238000004590 computer program Methods 0.000 title claims description 10
• 238000012512 characterization method Methods 0.000 claims description 32
• 230000003595 spectral effect Effects 0.000 claims description 16
• 108091026890 Coding region Proteins 0.000 claims description 10
• 238000012545 processing Methods 0.000 claims description 9
• 230000005236 sound signal Effects 0.000 claims description 7
• 238000001228 spectrum Methods 0.000 claims description 4
• PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 claims 1
• 108700026244 Open Reading Frames Proteins 0.000 claims 1
• 230000003044 adaptive effect Effects 0.000 description 10
• 238000005516 engineering process Methods 0.000 description 8
• 230000005540 biological transmission Effects 0.000 description 6
• 230000008901 benefit Effects 0.000 description 4
• 238000010586 diagram Methods 0.000 description 3
• 230000009977 dual effect Effects 0.000 description 3
• 238000005192 partition Methods 0.000 description 3
• 230000011218 segmentation Effects 0.000 description 3
• 238000004458 analytical method Methods 0.000 description 2
• 238000010276 construction Methods 0.000 description 2
• 238000002372 labelling Methods 0.000 description 2
• 238000013139 quantization Methods 0.000 description 2
• 238000005070 sampling Methods 0.000 description 2
• 101000969688 Homo sapiens Macrophage-expressed gene 1 protein Proteins 0.000 description 1
• 102100021285 Macrophage-expressed gene 1 protein Human genes 0.000 description 1
• 230000015572 biosynthetic process Effects 0.000 description 1
• 238000006243 chemical reaction Methods 0.000 description 1
• 238000006073 displacement reaction Methods 0.000 description 1
• 238000009826 distribution Methods 0.000 description 1
• 238000001914 filtration Methods 0.000 description 1
• 238000012986 modification Methods 0.000 description 1
• 230000004048 modification Effects 0.000 description 1
• 238000005457 optimization Methods 0.000 description 1
• 230000008447 perception Effects 0.000 description 1
• 230000010363 phase shift Effects 0.000 description 1
• 230000011664 signaling Effects 0.000 description 1
• 238000012360 testing method Methods 0.000 description 1
• 238000012546 transfer Methods 0.000 description 1

• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
  • H03M7/00—Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
  • H03M7/30—Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
  • H03M7/40—Conversion to or from variable length codes, e.g. Shannon-Fano code, Huffman code, Morse code
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
  • G10L19/008—Multichannel audio signal coding or decoding using interchannel correlation to reduce redundancy, e.g. joint-stereo, intensity-coding or matrixing
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
  • G10L19/04—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques
  • G10L19/08—Determination or coding of the excitation function; Determination or coding of the long-term prediction parameters
• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
  • H03M7/00—Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
  • H03M7/30—Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
  • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
  • H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
  • H04N19/12—Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
  • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
  • H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
  • H04N19/146—Data rate or code amount at the encoder output
  • H04N19/15—Data rate or code amount at the encoder output by monitoring actual compressed data size at the memory before deciding storage at the transmission buffer
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
  • H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
  • H04N19/189—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding
  • H04N19/196—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the adaptation method, adaptation tool or adaptation type used for the adaptive coding being specially adapted for the computation of encoding parameters, e.g. by averaging previously computed encoding parameters
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
  • H04N19/46—Embedding additional information in the video signal during the compression process
  • H04N19/463—Embedding additional information in the video signal during the compression process by compressing encoding parameters before transmission
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
  • H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
  • H04N19/90—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using coding techniques not provided for in groups H04N19/10-H04N19/85, e.g. fractals
  • H04N19/91—Entropy coding, e.g. variable length coding [VLC] or arithmetic coding
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
  • G10L19/0017—Lossless audio signal coding; Perfect reconstruction of coded audio signal by transmission of coding error
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
  • G10L19/002—Dynamic bit allocation
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L19/00—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
  • G10L19/02—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders
  • G10L19/0204—Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using spectral analysis, e.g. transform vocoders or subband vocoders using subband decomposition
• • G—PHYSICS
  • G10—MUSICAL INSTRUMENTS; ACOUSTICS
  • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
  • G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
  • G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
  • G10L25/18—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being spectral information of each sub-band

• Engineering & Computer Science (AREA)
• Multimedia (AREA)
• Signal Processing (AREA)
• Physics & Mathematics (AREA)
• Theoretical Computer Science (AREA)
• Computational Linguistics (AREA)
• Health & Medical Sciences (AREA)
• Audiology, Speech & Language Pathology (AREA)
• Human Computer Interaction (AREA)
• Acoustics & Sound (AREA)
• Discrete Mathematics (AREA)
• General Physics & Mathematics (AREA)
• Mathematical Physics (AREA)
• Computing Systems (AREA)
• Compression, Expansion, Code Conversion, And Decoders (AREA)

1321316 九、發明說明： 【發明所屬之技術領域】 本發明關於參數之無損編碼，且尤其是關於有效 數之編碼規則的產生和使用》 【先前技術】 近期’多聲道音訊再生技術變的更加重要。這可 貫上爲如熟知之MP3技術的音訊壓縮/編碼技術已 能透過網際網路或具有限頻寬之其他傳輸通道加以 訊紀錄。MP3編碼技術已變得如此知名，因事實上它 立體聲格式散佈所有記錄，該立體聲格式亦即爲含 左立體聲道及第二或右立體聲道之音訊紀錄之數位f 儘管如此，仍有習知兩聲道音響系統之基本缺點 ，已發展出環繞技術。除兩立體聲道L和R外，所 多聲道環繞呈現包含一額外之中央聲道C和兩個環 Ls、Rs。該參數音響格式亦稱爲三/雙立體聲，其 個前聲道和兩個環繞聲道。通常，需要五個傳輸聲 播放環境中，需要至少五個喇叭擺設在五個適當位 適當距離之五個良好擺設的喇叭中得到最佳之甜美ΐ 在降低傳輸多聲道音訊信號所需資料量之技術中 有數種技術。這種技術稱爲聯合立體聲技術。最後參 圖，其表示聯合立體聲裝置60»該裝置可爲實施， 強度立體聲（I S )或雙聲道佇列編碼（BCC )之裝置。這 通常接收至少兩個聲道（CHI、CH2、…CHn)作爲輸入 出至少一單載波聲道與參數資料。將該參數資料作 壓縮參 由於事 使其可 散佈音 允許以 第一或 6表。 。因此 建議之 繞聲道 意爲三 道。在 置，在 !點。 知道的 考第9 例如， 種裝置 ，並輸 爲界定 1321316 位使得在一解碼器中可計算原始聲道（CHI、CH2、…CHn)之 近似値。 正常上，載波聲道將包含次頻帶取樣、頻譜係數、時域 取樣等，其提供相當細緻之根本信號代表，而參數資料不 包含這種頻譜係數取樣，但包含用以控制如藉乘積、時間 位移、頻率位移 '相位位移等之權重的某一重建演算法之 控制參數。因此，以數字開始，載波聲道所需資料量範圍 爲60 - 70kbi t s/s，而一個聲道之參數側資訊所需資料量範 圍一項爲1.5-2.5kbits/s。如以下之說明’參數資料實例 爲熟知之比例因子，強度立體聲資訊或雙聲道佇列參數。 在2002年5月於慕尼黑“_1(：}1)，福勒（(：吓八1^1^1〇、巴 加德（F . BAUMGARTE)在AES會議文件5 5 74”應用至立體聲和 多聲道音訊壓縮之雙聲道佇列編碼”中，在200 1年10月於 紐約莫漢克（M0H0NK)，福勒和巴加德在IEEE WASPAA文件” 使用知覺參數化之空間音訊的有效呈現”、”應用至以彈性表 達之音訊壓縮的雙聲道佇列編碼”中，2002年10月於洛山 磯，113屆AES會議之未出版文件5686及2003年11月， 福勒和巴加德在第1 1冊，語音和音訊處理，I EEE電晶體之 ”雙聲道佇列編碼-第Π部：結構及應用”中例如說明了 BCC 技術。 在BCC編碼中，使用具有重疊型窗之DFT(離散式傅立葉 轉換）爲基準之轉換將多數音訊輸入聲道轉成頻譜表徵。所 形成之均勻頻譜分成非重疊型雙分區。各分區之頻寬大致 上與對等方型頻寬爲正比。然後在各個方區之兩個聲道之 1321316 間槪算BCC參數。各個聲道對於參數聲道正常上附有這些 BCC參數且還將其量化。依據規定公式（已編碼）最後加以計 算已傳輸之參數，該參數可亦依要處理訊號之特定分區而 定。 多數BCC參數真的存在》例如，ICLD參數說明在兩個比 較聲道中所含能量之差（比率）。I CC參數（聲道間一致性/ 相互關係）說明兩個聲道間之相互關係，可將其了解爲兩個 聲道波形之類似性。I CTD參數（聲道間時差）說明兩個聲道 之通用時間位移，但是I PD參數（聲道間相位差）針對信號 相位作相同之說明。 在音訊信號之訊框似處理中，我們應知道BCC分析亦實 施訊框似，及時間變化。以頻率似之分析。對於各頻語帶 ，這意爲個別取得BCC參數。如果音訊過濾庫將輸入信號 分解成，例如，3 2個帶通信號時，這更意爲BCC分析區對 各32個頻帶取得BCC參數集。 在2004年5月，布里巴特（J. BREEBAART )、克勞許（A . KOHLRAUSCH)、舒依特（SCHUIJERS)在柏林（BERLIN)第 1 16 屆AES會議之未出版文件6072中”低位率下之高品質參數 空間音訊編碼”中，及2004年5月’舒依特、巴里布特、 彭海根（PURNHAGEN)、英德加（ENGDEGARD)在柏林第1 16屆 AES會議之未出版文件6073 ”低複雜性參數立體聲編碼”中 說明亦名爲參數立體聲之相關技術。 總而言之，多聲道音訊信號之參數編碼的最近方法（"空 間音訊編碼"、"雙聲道佇列編碼”（BCC )等）利用下行混合（ 1321316 可能爲單音或含數個聲道）及具有感覺空間音感劇場之參 數側資訊（“空間佇列”）呈現多聲道音訊信號。爲了使虛耗資 訊降至最低並使下行混合信號之編碼有儘多可利用之傳輸 容量，預期儘可能保持低側資訊° 保持低側資訊位元率之一種方式爲對側資訊施加’例如 ，熵編碼演算法，將空間音訊結構之側資訊無損地加以編 碼。 爲確保量化頻譜係數和某它側資訊之最佳化緊密表現’ 已將無損編碼延伸地應用在一般音訊編碼中。在I SO / I EC 標準MPEG1第3部、MPEG2第7部和MPEG4第3部當中有適 當編碼結構和方法之實例。 這些標準以及例如，1 99 7年 1〇月奎肯布希（S.R. QUACKENBUSH)、強斯頓（J · D . JOHNSTON)在紐約莫漢克 IEEE 之IEE文件”MPEG-2高等音訊編碼中，量化頻譜係數之無雜 訊編碼”說明包含以下測量無損編碼量化參數之現代科技： 鲁量化頻譜係數之多維霍夫曼編碼 *對係數集使用普遍（多維）之霍夫曼編碼簿 *將該職編碼成孔或分別爲編碼符號資訊與大小資訊（ 即’對於降低必要編碼簿大小，”有符號”對，，無符號” 編碼簿之特定絕對値僅具霍夫曼編碼簿項次） *在要編碼參數當中使用不同最大絕對値（lav)，即，不 同最大絕對値之另選邊編碼簿。 *對各LAV使用不同統計分佈之另選編碼簿 *傳輸選取霍夫曼編碼簿作爲側資訊給解碼器 1321316 *使用”分段”界定各選取之霍夫曼編碼簿的應用範圍 •針對頻率和隨後霍夫曼編碼結果之差分比例因子編碼 在MPEG 1音頻標準當中提議將粗略量化直無損失地編碼 成單一 PCM之另一種技術（在該標準中稱爲分群並供第2 層用）。這在標準ISO/IEC 1172-3:93中有更詳細說明。 福勒和巴加德在2003年11月第11冊，第6號之有關語 音和音訊處理的IEEE電晶體”雙聲道佇列編碼-第E部： 結構和應用”之出版給予有關BCC參數編碼之某些資訊。 提議的事將量化I CLD參數差別性地針對以下加以編碼 *針對頻率且（以一維霍夫曼碼）將該結果隨後施以霍夫 曼編碼 ♦針對時間且（以一維霍夫曼碼）將該結果隨後施以霍夫 曼編碼 且最後，選許更多有效變種，代表原始音訊信號。 如以上所提及，已經提議針對頻率且另外針對時間加差 分編碼使壓縮效能達到最佳化，並選取更多有效之變種 。然後經由某些側資訊將已選取之變種發訊至解碼器。 已經很花某些心力在降低下行混何音訊聲道與相對應側 資訊之大小。儘管如此，給予每一可能應用之可達成位 元率仍太高。例如，對行動電話之音訊和視訊內容串流 需要最不可能之位元率且因此對該內容有更有效之編碼 〇 【發明內容】 本發明之目的在提供一種增進編碼槪念’以較高效能達 1321316 成參數値之無損壓縮》 依據本發明之第一觀點，本目的之達成世界參數壓縮用 之壓縮單元，該參數包含代表原始信號之第一部位的第一 參數集，該參數更包含代表該原始信號之第二部參數集， 該第二部毗鄰該第一部位，包含：用於供給第一元組（TUPLE) 和第二元組之供給器，各元組具有至少兩個參數，該第一 元組之兩個參數來自第一參數集而第二元組之一個參數來 自第一參數集且一個參數來自第二參數；位元估算器，係 依據編碼規則用以估算使用含第一元組之第一元組序列將 參數集加以編碼，以及使用含第二元組之第二元組序列將 參數集加以編碼所需之位元數；以及用於供應已編碼區之 供應器，該供應器可作用在使用形成較低位元數之元組序 列’供應已編碼區，並用於供應表示元組序列之序列標示 ’其中之編碼區係得自該元組序列。 依據本發明之第二觀點，本目的之達成世界用於將已編 碼之參數區加以解碼之解碼器，該參數包含代表原始信號 之第一部位的第二參數集，該第二部位毗鄰該第一部位， 並用於處理序列標示，包含：解壓縮器，該解壓縮器可作 用爲使用以將元組序列編碼之編碼規則，將已編碼之參數 ^力Π以解壓縮而得到參數之元組序列，各該元組具有至少 Μ 參數；以及用於接收序列標是知訊框建構器，該序列 標示表示來自已編碼區下方之多數不同序列的已用過元組 $列’並使用以用過之元組列的資訊加以建構參數集。 依據本發明之第三觀點，本目的之達成世界參數壓縮用 -10- 1321316 之方法，該參數包含代表原始信號之第一部位 集，該參數更包含代表原始信號之第一部位的 ，該參數更包含代表原始信號之第二部位的第 該第二部位毗鄰第一部位。 依據本發明之第四觀點，本目的之達成當在 時，實施以上方法之電腦程式。 依據本發明之第五觀點，本目的之達成是藉 區解碼用之方法，該參數更包含代表原始信號 的第二參數集，該第二部位毗鄰該第一部位， 序列標示。 依據本發明之第六觀點，本目的之達成是藉 執行時，實施以上方法之電腦程式。 依據本發明之第七觀點，本目的之達成是藉 表徵，該參數包含代表原始信號之第一部位的 ，該參數更包含代表原始信號之第二部位的第 該第二部位毗鄰原始信號之第一部位，包含： 元組件序列之已編碼參數區下方之第一或第二 過元組序列，其中該第一序列包含有兩個參數 參數集：第一元組且其中該第二序列包含有一 自第一參數集而一個參數是來自第二參數集之） 本發明是依據發現到當將參數佈置在第一元 二元組序列中時，可將包含代表原始信號第一 參數集和包含代表原始信號第二部位之第二參 有效地加以編碼，其中該第一元組序列包含有 的第一參數 第一參數集 二參數集， 電腦上執行 已編碼參數 之第一部位 並用於處理 當在電腦上 參數之壓縮 第一參數集 二參數集， 代表已用過 序列的已用 是來自第一 個參數是來 g二元組^ 組序列和第 部位之第一 數集的參數 兩個參數是 -11- 1321316 來自原始信號單一埠位之參數元組且其中該第二元組序列 包含有一個參數是來自原始信號第一部位而一個參數是來 自原始信號第二部位之參數元組。有效的編碼之達成可使 用估算將第一和第二元組序列加以編碼所需位元數之位元 估算器，其中只將元組序列加以編碼，其形成低位元數。 因此，基本原理爲重新佈置要編碼之參數，例如爲時間 和頻率，且最後使用參數之一種佈置（元組序列）供壓縮， 其造成壓縮參數之低位元數。 在本發明之一實施例中，將說明代表原始信號之兩連續 時間部位的頻譜之兩頻譜參數集調適地分群爲兩對參數， 增強編碼效能。因此，一方面使用由來自相同時間部之兩 毗鄰頻率參數所組成之參數元組，產生元組序列。另一方 面，使用以來自第一時間部之第一參數和來自原始信號之 第二時間部的對應參數所建構：元組，產生第二元組序列 。然後，使用二維霍夫曼碼將兩個元組序列加以編碼。比 較兩個已編碼之元組序列大小且最後選取造成低位元數之 元組加以傳輸。將已使用.該種類元組加以建構已編碼資料 之資訊傳輸至解碼器作爲附加側資訊》 由於將參數紛尋成由兩參數所組成之元組，先前說明之 發明性編碼器的一項優點爲可將二維霍夫曼碼應用於壓縮 ，其通常導致低位元率。 第二優點爲，於編碼程序期間在兩個可能分群策略之間 作動態決定之調適型分群槪念產生另降低側資訊之位元率 -12- 1321316 對於兩個連續訊框集，在兩分群策略之間決定僅一次’ 府加降低所需側資訊量，因對於兩個完整連續時框集只需 將在編碼期間以利用分群策略之標示傳輸一次。 在本發明另一個實施例中。發明性之壓縮單元附加包含 在調適型分群前以時間或頻率將參數加以區分編碼之差分 編碼器。該差分編碼以及調適型分群和適當之霍夫曼編碼 簿更降低側資訊大小加以傳輸。兩種差分編碼之可能性以 及兩個分群策略造成總共四種可能組合。更增加發現編碼 規則之槪率，其造成低側資訊之位元率。 在本發明另一個實施例中，爲壓縮單元使用槪念，根據 使以編碼：參數區下方的分群結構發出信號之側資訊，允 許將已編碼之參數區解碼並重建原使訊框。在有利修飾中 ，該發明性解碼器亦允許尙未調適性地加以分群之資訊解 碼，因此，可達成該發明性解碼器與現有設備之相容性。 【實施方式】 第1圖表是發明性壓縮單元90，包含供給器100，位元 估算器102及供應器104。 供給器100在兩資料輸出處供給第一元組序列l〇6a和第 二元組序列1 0 6 b。供應器1 〇 4在其兩個資料輸出1 〇 8 a和 l〇8b上接收元組106a和106b。位元估算器在其資料輸入 1 10a和1 10b上接收兩位元組。 位元估算器1 0 2估算由對兩元組1 06 a和1 〇 6 b施加編碼 規則所形成之位元數。不管元組1 0 6 a或1 〇 6 b是否會造成 低位元數’位元估算器102選取造成低位元數之元組及透 -13- 1321316 過發訊輸出112a之訊號。 依據位元估算器102之決定，將形成低位元數之元組最 後編碼區118 '該編碼區是透過供應器104之輸出120a佳 以提供，其中，供應器更在其發訊線1 2 〇 b發出序列標示訊 號’表示將原始元組（1 0 6 a或1 0 6 b )序列編碼，得到編碼區 118° 在另一實施例中，當省略供給器1 00和供應器1 04之間 的虛線連結1 2 2 a和1 2 2 b時，可達成相同功能性。在該替 代方案中’位元估算器1 〇 2會將元組1 0 6 a和1 0 6 b序列加 以編碼’且會傳送兩個相異編碼區i 24a和124b給供應器 1 04 ’其中’供應器附加發出訊號從該訊號得到原始元組 106a和l〇6b序列。已編碼區124 a和124b或可由供應器 104案中發出訊號》 在這另選實施例中，供應器丨〇 4會就只轉送低位元數之 編碼區到其輸出1 2 〇 a，附加提供序列標示。 第2圖表示兩個調適型分群結構之實例，使用該結構得 到編碼之元組序列。爲了說明發明性之調適型分群原理， 表示原始信號之四個後續時框13〇3至130d，其中，各時段 爲五頻譜參數132a至I32e集。 依據本發明’如四元組13421和i34b之說明，以頻率， 如位元組136a和136b之說明，以時間將兩連續時框之頻 譜參數加以分群’建立元組序列。時間分群形成第一元組 1 3 8序列’但是頻率分群形成第二元組M〇序列。 使用’例如’霍夫曼編碼簿將元組丨38和丨4〇序列加以 -14- 1321316 編碼，形成兩不同序列之碼字元142和144。依據本發明， 將需要更少位元數之碼字元序列最後傳輸至解碼器，不管 是否以時間分群或頻率分群爲碼字元序列之基礎發出信號 ，該解碼器附加接收一序列標示。如在第2圖中所能看到 的，對於所是之參數對（二維）的調適型分群實例，該序列 標示只含一單位元。 第3圖表示用以實施發明性調適分群之某些另選分群策 略，允許大於二維之霍夫曼碼。 第3圖表不二維霍夫曼碼146a，三維霍夫曼碼146b和四 維霍夫曼碼146 c之分群策略。對於各種策略，說明兩個連 續時框’其中，以相同大寫字母代表屬於相同元組之參數 〇 如在第2圖中已經說明的，完成二維霍夫曼碼情況之分 群’以頻率148a和時間148b建構二維元組。在建構含三 個參數之元組的情況中，頻率元組158a爲使在一個頻框內 之三個毗鄰頻率參數分群在一起，形成一元組。如在第3 圖中所示’將時間元組150b建構成將來自一時框之兩個毗 鄰參數和來自另一時框之一個參數組合在一起。 藉由將一時框之四個毗鄰參數分群爲一個元組，對應於 其他時間元組，建構四維時間分群爲一個元組1 5 2 a »將時 間分群元組152b建構成一時框之兩個毗鄰參數和另一時框 之兩個毗鄰參數組合在一起，其中單一時框之參數對說明 兩個連續時框之相同頻譜特性。 如第3圖中所示，允許不同分群結構可明顯降低側資訊 -15- 1321316 之位元率，例如，如果使用各種預先界定之不同維度的霍 夫曼編碼簿，可在編碼程序內改變分群維度，使得在編碼 程序內之任何時間可使用形成最低位元率之表徵。 第4圖表示如何使用附加包含差分編碼器之發明性壓縮 單元，在霍夫曼編碼程序前藉應用某種差分編碼進而降低 側資訊。 爲了以時間和頻率或以時間或以頻率說明差分編碼，使 用第2圖中表示之參數160的相同絕對表徵作爲各種差分 編碼步驟之基礎。第一種可能性爲以頻率將絕對參數16。 加以差分編碼，形成差分編碼參數162。如在第4圖中可看 到的，爲了將絕對表徵160家以差分編碼，使各時框之第 一參數不變，但是以絕對表徵1 60之第二參數和第一參數 的差取代第二參數。依照相同規則建構差分編碼表徵內之 其他參數。 另一種可能性爲以時間之差分編碼，產生表徵1 64。該表 徵之建構是使整個第一時框不變，但是，如第4圖中所能 見到的，以絕對表徵參數和先前時框相同參數之差取代下 —時框參數。 第三種可能性爲以頻率首先加以差分編碼，接著以時間 或反之加以差分編碼，兩種形成以時間和頻率所差分編碼 之相同編碼表徵166。 注意到有機會使用代表原始信號之那些四個信號作爲調 適型分群之輸入。看一下指定參數實例之表徵160至166 能清楚見到差分編碼對於側資訊之傳輸率有多大影響。看 -16- 1321316 —下絕對表徵1 6 Ο，認知到時間分群不會，頻率分群亦不會 形成相同內容之元組°因此無法建構會指定最短自原碼給 最常發生之元組的適當霍夫曼編碼簿。 該情況相異看到頻率差分編碼表徵162,當中可能建構只 需具有四個項次，涵蓋完全表徵之霍夫曼編碼簿’且當中 會指定元組（1 ’ 1 )或元組（2 ’ 2 )爲最小長度之碼字元’達 成精簡側資訊。 在時間差分編碼表徵1 6 4中該優點較不明顯。然而此處 亦能得到頻率分群及利用多個元組（5，5 )及（1 〇，1 〇 )。 就時間及頻率之差分編碼表徵166比起表徵162,甚至會 達成降低側資訊之位元率，因如圖中所示，時間分群會形 成高倍數之元組（1，0 )，允許建構會指定最短碼字元給先 前元組之霍夫曼編碼簿。 如清楚在第4圖中看到的，利用調適型分群和差分編碼 之發明槪念的高度彈性允許選取最適合原始音訊之策略， 因此允許保持低側資訊位元率。 總之’在一較佳實施例中’首先針對時間（變數i )和頻率 (變數2)差分編碼出量化參數値。然後將鎖形成之參數針對 時間（變數a )和頻率（變數b )加以調適分群。結果，有四種 組合（la’ lb’ 2a’ 2b)’從中選取最佳者並對解碼器發出 信號。這可能只以代表，例如，位元組合〇〇，〇 1，1 〇，！ i 變數la，lb，2a，2b的二位元資訊來完成。 第5圖表示依據目前發明之解碼器將已編碼之參數區加 以解碼，其中’該參數區包含具有第一頻譜參數集之第一 -17- 1321316 訊框集具有第二頻譜參數集之第二訊框。 解碼器200包含解壓縮器202與訊框建構器204。解壓縮 器在輸入上接收已編碼參數區206得到參數元組序列208 。將該參數元組序列208輸入至訊框建構器204。 訊框建構器附加接收序列標示210,表示編碼器已使用什 麼元組序列去建構已編碼之參數區。 然後訊框建構器204重排由序列標示210所引導之元組 序列208，從參數元組序列208重建第一訊框1 12a和第二 訊框1 12b。 上述之本發明較佳實施例藉由使用多維霍夫曼碼，導入 要編碼之値的調適型分群達成編碼效能之進一步增強。做 爲一實例，可完成針對頻率之二維分群値以及針對時間之 二維分群値。然後編碼結構會完成兩種型式之編碼並選取 更有利之一種（即，需較少位元之變種）。透過側資訊發出 該決定信號給解碼器。 如第3圖中之說明，在另一個實例中，亦可能建構更高 維度之霍夫曼碼，應用不同分群策略加以建構元組。該假 定實例表示只將來自兩個連續訊框之參數分群在一起，建 構元組之分群策略。亦可能利用來自三個或更多各連續訊 框之參數加以分群，以直接方式加以分群。 在發明性編碼器之修訂中，亦可能使用不同霍夫曼編碼 簿，組合差分分群和差分編碼策略得到最短可能之側資訊 表徵。這可能以具有附加側資訊參數爲代價，進而降低以 編碼音訊之側資訊位元率，發出編碼所用之霍夫曼編碼簿 -18- 1321316 信號。 本發明所說明之較佳實施例表示例如，發明性槪念，當 中分群策略在兩個連續時框內未改變。在本發明之修訂中 ，當然亦可能在兩個訊框及內之時間和頻率分群之間具多 重改變，其意爲在訊框之內亦供給序列標示，發出分群策 略變更之信號。 在假定實例中，於霍夫曼編碼前將該參數加以差分編碼 。在參數之霍夫曼編碼前當然每一其它無損編碼規則亦爲 可能’編碼之目標爲儘可能得到相同內容之元組。 第4圖中有四種不同之參數表徵，亦即，絕對表徵、頻 率差分表徵、時間差分表徵以及時間和頻率差分表徵。在 四個表徵中選擇，如第四圖所示，對已使用之表徵發出信 號之側資訊必須至少爲兩位元。爲了平衡贏得針對附加頻 譜表徵標示之可能的編碼效能效益，當然亦可能決定原則 上只允許兩個可能表徵，降低頻譜表徵表示爲單一位元長 度。 作爲發明性解碼器之實例，第5圖表示除以編碼參數區 206外’附加接收某些側資訊之解碼器200。在假定實例中 ’引導訊框建構器204當然可處理所需之任何其他側資訊 ’特別是頻譜表徵表示，表示已被使用來對原如訊框編碼 之頻譜表徵。 依發明方法之某些實施要求而定，可以硬體或軟體實施 本發明方法。該實施可使用數位儲存有電子可讀取控制信 號之磁碟，DVD或CD ’騎與可程式化之電腦系統合作，以 -19- 1321316 致能執行本發明方法。通常 在機器可讀取載具之電腦程 程式產品時，該程式碼作用 本發明方法因此爲電腦程式 執行該電腦程式時用以實施 0 雖然參數其特定實施例已 ，但是那些嫻熟技術者會了 其它變更而不偏離其精神和 施例做各種變更而不偏離此 含之廣義槪念。 【圖式簡單說明】 藉由參數附圖，隨後說明 第1圖表是發明性之壓縮 第2圖表示兩種可能性之 第3圖表是額外可能性之 第 4圖 表 是 差 分 編 碼 結 構 第 5圖 表 是 發 明 性 解 碼 器 第 6圖 表 是 先 -v /— 刖 技 術 之 多 【主要 部分之代表长 f號說明】 60 聯 合 -_jL- 體 聲 裝 置 90 壓 縮 單 元 100 供給器 本發明因此爲將程式碼儲存 產品，當在電腦上執行電腦 實施本發明方法。另言之， 該電腦程式具有當在電腦上 少其中一發明方法之程式碼 對前項特別加以表示與說明 到可在形式和細節上做各種 圍。要了解到可配合不同實 所發表及以下申請項目所包 發明之較佳實施例，其中： 元： 適型分群參數； 適型分群參數； 道編碼器 13213161321316 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to lossless coding of parameters, and in particular to the generation and use of coding rules for significant numbers. [Prior Art] Recently, the multi-channel audio reproduction technology has become more important. This is consistent with the fact that audio compression/encoding techniques such as the well-known MP3 technology have been able to record via the Internet or other transmission channels with limited bandwidth. MP3 encoding technology has become so well known, because in fact its stereo format spreads all the records, which is the digits of the audio record containing the left stereo channel and the second or right stereo channel. However, there are still two known The basic shortcomings of the channel sound system have developed surround technology. In addition to the two stereo channels L and R, the multi-channel surround presentation includes an additional center channel C and two rings Ls, Rs. The audio format of this parameter is also known as three/dual stereo, with its front channel and two surround channels. Generally, in a five-transmission sound playback environment, at least five speakers are required to be placed in five well-positioned speakers at the appropriate distances to obtain the best sweetness. The amount of data required to transmit multi-channel audio signals is reduced. There are several techniques in the technology. This technique is called joint stereo technology. Finally, reference is made to the Joint Stereo Device 60» which may be a device that implements Intensity Stereo (I S ) or Dual Channel Coding (BCC). This typically receives at least two channels (CHI, CH2, ... CHn) as input to at least one single carrier channel and parameter data. The parameter data is used as a compression parameter to make it possible to distribute the sound to the first or sixth table. . Therefore, the recommended surround sound means three channels. At the set, at the ! point. Knowing the test, for example, the device, and inputting the 1321316 bits, makes it possible to calculate the approximate 値 of the original channels (CHI, CH2, ... CHn) in a decoder. Normally, the carrier channel will contain sub-band sampling, spectral coefficients, time-domain sampling, etc., which provide a fairly detailed representation of the fundamental signal, while the parametric data does not contain such spectral coefficient samples, but is included to control such as multiplication, time, etc. The control parameters of a reconstruction algorithm for the weight of the displacement, frequency shift 'phase shift, etc. Therefore, starting with a number, the amount of data required for the carrier channel is 60 - 70kbi t s/s, and the amount of data required for the parameter side information of one channel is 1.5-2.5kbits/s. As explained below, the parameter data examples are well-known scale factors, intensity stereo information or binaural parameters. In May 2002 in Munich "_1 (:} 1), Fowler ((: scared eight 1 ^ 1 ^ 1 〇, Bagard (F. BAUMGARTE) in the AES meeting document 5 5 74" applied to stereo and more In the two-channel array coding of channel audio compression, in October 2001 in New York, Mohank (M0H0NK), Fowler and Bagard in the IEEE WASPAA file "using the spatial representation of perceptually parameterized spatial audio "," applied to the two-channel array coding of audio compression with elasticity," in October 2002 in Los Angeles, the unpublished document of the 113th AES Conference, 5686 and November 2003, Fowler and Baga In the first volume, voice and audio processing, I EEE transistor "two-channel 编码 column coding - the third part: structure and application", for example, the BCC technology is described. In BCC coding, the use of overlapping windows The DFT (Discrete Fourier Transform)-based conversion converts most audio input channels into spectral representations. The resulting uniform spectrum is divided into non-overlapping dual partitions. The bandwidth of each partition is roughly equal to the equivalent square bandwidth. Proportional. Then calculate the BCC parameters between 1321316 of the two channels in each square. Each channel is normally accompanied by these BCC parameters for the parameter channel and is also quantized. The transmitted parameters are finally calculated according to the prescribed formula (encoded), which may also depend on the particular partition of the signal to be processed. The BCC parameter really exists. For example, the ICLD parameter specifies the difference (ratio) of the energy contained in the two comparison channels. The I CC parameter (inter-channel consistency/relationship) illustrates the relationship between the two channels, It can be understood as the similarity of the two channel waveforms. I CTD parameters (inter-channel time difference) illustrate the universal time shift of the two channels, but the I PD parameters (inter-channel phase difference) are the same for the signal phase. In the frame-like processing of audio signals, we should know that BCC analysis also implements frame-like and time-varying. It is analyzed by frequency. For each frequency band, this means to obtain BCC parameters individually. If audio filtering The library decomposes the input signal into, for example, 32 2 band-pass signals, which means that the BCC analysis area obtains the BCC parameter set for each of the 32 frequency bands. In May 2004, Bribart (J. BREEBAART), gram A. KOHLRAUSCH and SCHUIJERS in the unpublished document 6072 of the 1st AES Conference in Berlin (BERLIN), "High Quality Parameter Space Audio Coding at Low Rate", and May 2004 'Suit, Barry Butte, PURNHAGEN, ENGDEGARD at the 16th AES Conference in Berlin, unpublished document 6073, "Low Complexity Parameter Stereo Coding", also describes the parametric stereo correlation In summary, the most recent method of parameter encoding for multi-channel audio signals ("spatial audio coding", "two-channel constellation coding (BCC), etc.) utilizes downmixing (1321316 may be mono or alpha) The channel side information and the parameter side information ("space array") with the sense space sound perception theater present multi-channel audio signals. In order to minimize the lossy information and to make the downlink mixed signal code have as much available transmission capacity, it is expected to keep the low side information as much as possible. One way to keep the low side information bit rate is to apply 'for the opposite side information', for example, The entropy coding algorithm encodes the side information of the spatial audio structure without loss. To ensure that the quantized spectral coefficients and the optimization of some of its information are closely represented, lossless coding has been extended to general audio coding. Examples of appropriate coding structures and methods are found in I SO / I EC Standard MPEG1 Part 3, MPEG 2 Part 7, and MPEG4 Part 3. These standards are, for example, quantified by SR QUACKENBUSH and J.D. JOHNSTON in the IEE document "MPEG-2 Advanced Audio Coding" of New York Mohank IEEE in 1978. The “no-noise coding of spectral coefficients” description contains the following modern techniques for measuring lossless coding quantization parameters: Multi-dimensional Huffman coding of Lu-quantized spectral coefficients* Using a universal (multi-dimensional) Huffman codebook for coefficient sets* Holes are either coded symbol information and size information (ie 'for reduced necessary codebook size, 'signed) pairs, unsigned" codebook specific absolute 値 only Huffman codebook entries) The coding parameters use different maximum absolute lav (lav), that is, different maximum absolute 値 alternative side codebooks. * Use a different codebook for each LAV with different statistical distributions. * Transfer the Huffman codebook as side information. Decoder 1321316 * Use "segmentation" to define the application range of each selected Huffman codebook • Differential scale factor encoding for frequency and subsequent Huffman coding results in MPEG 1 audio Another technique for encoding coarsely quantized straight loss without loss into a single PCM (referred to as grouping and for layer 2 in this standard) is proposed in more detail in the standard ISO/IEC 1172-3:93. Fowler and Bagard in the November 2003, Volume 11, No. 6 IEEE Speech on Audio and Audio Processing "Double-Channel Parallel Coding - Part E: Structures and Applications" published on BCC parameters Some information about the encoding. The proposed thing will quantify the I CLD parameters differentially for the following * for the frequency and (with a one-dimensional Huffman code) then the result is then Huffman coding ♦ for time and The one-dimensional Huffman code) then applies Huffman coding to the result and finally, selects more valid variants, representing the original audio signal. As mentioned above, it has been proposed for frequency and additionally for time plus differential coding. The compression performance is optimized, and more effective variants are selected. Then the selected variants are sent to the decoder via some side information. It has taken some effort to reduce the downlink audio channel and the corresponding side. News Size. However, the achievable bit rate for each possible application is still too high. For example, the audio and video content streams for mobile phones require the least significant bit rate and therefore a more efficient encoding of the content. 〇 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 增进 增进 增进 增进 增进 增进 增进 增进 增进 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 132 The parameter includes a first parameter set representing a first portion of the original signal, the parameter further comprising a second parameter set representing the original signal, the second portion being adjacent to the first portion, comprising: for supplying the first tuple (TUPLE) and the second tuple feeder, each tuple has at least two parameters, two parameters of the first tuple are from the first parameter set and one parameter of the second tuple is from the first parameter set and one The parameter is from the second parameter; the bit estimator is used to estimate the encoding of the parameter set using the first tuple sequence containing the first tuple according to the encoding rule, and the second parameter is used The second tuple sequence of the tuple encodes the number of bits required to encode the parameter set; and the provider for supplying the encoded region, the provider can act on the use of a tuple sequence forming a lower number of bits The coded region is used to supply a sequence designation indicating a sequence of tuples, wherein the coding region is derived from the sequence of the tuple. According to a second aspect of the present invention, the object of the present invention is a decoder for decoding an encoded parameter region, the parameter comprising a second parameter set representing a first portion of the original signal, the second portion being adjacent to the first a portion, and for processing sequence identifiers, comprising: a decompressor, the decompressor can be used as an encoding rule for encoding a sequence of tuples, and the encoded parameters are decompressed to obtain a tuple of parameters a sequence, each of the tuples having at least a Μ parameter; and a receiving sequence indicator is a knowledge frame constructor that indicates a used tuple $column from a plurality of different sequences below the encoded region and is used The information in the tuple column is constructed to construct a parameter set. According to a third aspect of the present invention, the method of the present invention achieves the world parameter compression -10- 1321316, the parameter comprising a first part set representing the original signal, the parameter further comprising a first part representing the original signal, the parameter The second portion, further comprising a second portion representing the original signal, is adjacent to the first portion. According to a fourth aspect of the present invention, the object of the present invention is to implement a computer program of the above method. According to a fifth aspect of the present invention, the object is achieved by a method for borrowing a region, the parameter further comprising a second parameter set representing the original signal, the second portion being adjacent to the first portion, the sequence being labeled. According to a sixth aspect of the present invention, the object of the present invention is to implement a computer program of the above method at the time of execution. According to a seventh aspect of the present invention, the object is achieved by a representation comprising a first portion representing an original signal, the parameter further comprising a second portion of the second portion representing the original signal adjacent to the original signal a portion comprising: a first or second sub-tuple sequence below the encoded parameter region of the meta-component sequence, wherein the first sequence includes two parameter parameter sets: a first tuple and wherein the second sequence includes The first parameter set and one parameter is from the second parameter set. The invention is based on the discovery that when the parameters are arranged in the first meta-binary sequence, the first parameter set and the representative representative representing the original signal can be included The second parameter of the second portion of the original signal is effectively encoded, wherein the first tuple sequence includes a first parameter first parameter set two parameter set, and the first part of the encoded parameter is executed on the computer and used for processing The parameter on the computer is compressed by the first parameter set two parameter set, which represents the used sequence is used from the first parameter is to g binary group ^ group sequence and the first part The parameters of the first set of bits are two parameters: -11-1321316, a parameter tuple from a single unit of the original signal and wherein the second tuple sequence contains one parameter from the first part of the original signal and one parameter is from The parameter tuple of the second part of the original signal. The achievement of a valid code can use a bit estimator that estimates the number of bits needed to encode the first and second tuple sequences, wherein only the tuple sequence is encoded, which forms a low order number. Therefore, the basic principle is to rearrange the parameters to be encoded, such as time and frequency, and finally use one of the parameters (tuple sequence) for compression, which results in the lower number of bits of the compression parameter. In one embodiment of the present invention, two spectral parameter sets representing the spectrum of two consecutive time portions of the original signal are adaptively grouped into two pairs of parameters to enhance coding efficiency. Thus, on the one hand, a tuple sequence is generated using parameter tuples consisting of two adjacent frequency parameters from the same time portion. On the other hand, a second tuple sequence is generated using a tuple constructed from a first parameter from the first time portion and a corresponding parameter from the second time portion of the original signal. The two tuple sequences are then encoded using a two-dimensional Huffman code. A tuple that compares the size of the two encoded tuple sequences and finally selects the low-order number is transmitted. The information of the type of tuple used to construct the encoded data is transmitted to the decoder as additional side information. Since the parameters are searched for a tuple composed of two parameters, an advantage of the previously described inventive encoder is described. In order to apply a two-dimensional Huffman code to compression, it typically results in a low bit rate. The second advantage is that the adaptive grouping mourning that dynamically determines between two possible grouping strategies during the encoding process produces a bit rate of the lower side information -12-1321316 for two consecutive frame sets, in two groups The decision between the policies is only one time to reduce the amount of information required on the side, since only two consecutive consecutive time frame sets will need to be transmitted once during the coding period using the labeling strategy. In another embodiment of the invention. The inventive compression unit additionally includes a differential encoder that encodes the parameters in a time or frequency prior to the adaptive grouping. The differential coding as well as the adaptive segmentation and the appropriate Huffman codebook reduce the side information size for transmission. The possibility of two differential encodings and two grouping strategies result in a total of four possible combinations. The rate of discovery of the encoding rule is increased, which results in the bit rate of the low side information. In another embodiment of the present invention, the use of the commemoration for the compression unit allows the encoded parameter region to be decoded and reconstructed from the original frame based on the side information that signals the packetized structure below the parameter region. In an advantageous modification, the inventive decoder also allows for uncoordinated grouping of information decoding, thus achieving compatibility of the inventive decoder with existing equipment. [Embodiment] The first graph is an inventive compression unit 90 including a feeder 100, a bit estimator 102, and a supplier 104. The feeder 100 supplies a first tuple sequence l〇6a and a second binary sequence 1 0 6 b at the two data outputs. The provider 1 〇 4 receives the tuples 106a and 106b on its two data outputs 1 〇 8 a and l 〇 8b. The bit estimator receives a two-tuple on its data inputs 1 10a and 1 10b. The bit estimator 1 0 2 estimates the number of bits formed by applying encoding rules to the two sets of 06a and 1 〇 6 b. Whether or not the tuple 1 0 6 a or 1 〇 6 b causes the low byte number 'bit estimator 102 to select the tuple resulting in the lower truncation number and the signal through the -13-1321316 over-issue output 112a. Based on the decision of the bit estimator 102, the tuple last code region 118' which forms the low byte number is provided by the output 120a of the provider 104, wherein the provider is at its transmission line 1 2 〇 b issuing a sequence designation signal ' indicates that the original tuple (1 0 6 a or 1 0 6 b ) sequence is encoded to obtain a coding region 118°. In another embodiment, when the supplier 100 and the supplier 104 are omitted When the dotted line connects 1 2 2 a and 1 2 2 b, the same functionality can be achieved. In this alternative, the bit estimator 1 〇2 encodes the tuples 1 0 6 a and 1 0 6 b sequences and transmits two distinct coding regions i 24a and 124b to the supplier 104 ' The 'supplier appends a signal to obtain a sequence of original tuples 106a and 106b from the signal. The coded areas 124a and 124b may be signaled by the provider 104. In this alternative embodiment, the provider 丨〇4 will only forward the coded area of the lower number of bits to its output 1 2 〇a, additionally provided Sequence labeling. Figure 2 shows an example of two adaptive clustering structures that are used to obtain a sequence of encoded tuples. To illustrate the inventive adaptive clustering principle, four subsequent time frames 13 〇 3 to 130 d of the original signal are represented, wherein each time period is a set of five spectral parameters 132a through I32e. In accordance with the present invention' as illustrated by quads 13421 and i34b, the sequence parameters of the two consecutive time frames are grouped' by time, as illustrated by the bits 136a and 136b, to create a tuple sequence. Time grouping forms a first tuple 1 3 8 sequence ' but frequency grouping forms a second tuple M 〇 sequence. The tuples 丨38 and 丨4〇 sequences are encoded with -14-1321316 using a 'for example' Huffman codebook to form two different sequences of code characters 142 and 144. In accordance with the present invention, a sequence of codewords requiring fewer bits is ultimately transmitted to the decoder, and the decoder additionally receives a sequence of indications, whether or not signals are signaled on a time-sequence or frequency-segment basis. As can be seen in Figure 2, for a paired (two-dimensional) instance of the parameter pair, the sequence label contains only one unit. Figure 3 shows some alternative clustering strategies for implementing inventive adaptive clustering, allowing Huffman codes greater than two dimensions. The third chart is a grouping strategy of the two-dimensional Huffman code 146a, the three-dimensional Huffman code 146b, and the four-dimensional Huffman code 146c. For various strategies, two consecutive time frames are illustrated, in which the same uppercase letters represent parameters belonging to the same tuple, as already explained in Figure 2, completing the grouping of the two-dimensional Huffman code case by frequency 148a And time 148b constructs a two-dimensional tuple. In the case of constructing a tuple containing three parameters, the frequency tuple 158a forms a tuple for grouping together three adjacent frequency parameters within one frequency bin. As shown in Fig. 3, the time tuple 150b is constructed to combine two neighboring parameters from a one-time frame with one parameter from another time frame. By grouping the four adjacent parameters of the one-time frame into one tuple, corresponding to other time tuples, constructing a four-dimensional time grouping into one tuple 1 5 2 a » constructing the time grouping tuple 152b into two adjacent frames The parameter is combined with two adjacent parameters of the other time frame, wherein the parameter pair of the single time frame describes the same spectral characteristics of the two consecutive time frames. As shown in Figure 3, allowing different clustering structures can significantly reduce the bit rate of side information -15-1321316. For example, if various pre-defined different dimensions of the Huffman codebook are used, the grouping can be changed within the encoding program. The dimension is such that the representation that forms the lowest bit rate can be used at any time within the encoding process. Figure 4 shows how to use the inventive inventive compression unit with a differential encoder to apply some kind of differential coding before the Huffman coding procedure to reduce the side information. To account for differential encoding in time and frequency or in time or frequency, the same absolute representation of the parameter 160 represented in Figure 2 is used as the basis for the various differential encoding steps. The first possibility is to have an absolute parameter of 16 at frequency. Differential encoding is performed to form a differential encoding parameter 162. As can be seen in Fig. 4, in order to differentially encode 160 families, the first parameter of each time frame is unchanged, but the difference between the second parameter of the absolute representation 1 and the first parameter is substituted. Two parameters. Other parameters within the differential code representation are constructed according to the same rules. Another possibility is to encode in a time difference to produce a representation 1 64. The construction of the eigenstate is such that the entire first time frame is unchanged, but, as can be seen in Fig. 4, the lower-time box parameters are replaced by the difference between the absolute characterization parameters and the same parameters of the previous time frame. A third possibility is to first differentially encode the frequency, then differentially encode it in time or vice versa, and the two form the same coded representation 166 that is differentially encoded in time and frequency. Note the opportunity to use the four signals representing the original signal as input to the adaptive grouping. Looking at the characterization of the specified parameter instances 160 to 166, it is clear how differential encoding affects the transmission rate of the side information. See 16- 1321316 - the absolute representation of 16 Ο, knowing that time grouping will not, frequency grouping will not form the same content of the tuple. Therefore, it is impossible to construct the appropriate shortest from the original code to the most frequently occurring tuple. Hoffman codebook. This situation differs from the frequency differential coding representation 162, which may be constructed with only four terms, covering the fully characterized Huffman codebook' and in which a tuple (1 '1) or tuple (2' is specified. 2) Achieve reduced side information for the minimum length code character '. This advantage is less pronounced in the time differential coding characterization 146. However, frequency grouping and multi-tuples (5, 5) and (1 〇, 1 〇) can be obtained here. The differential coding representation 166 of time and frequency may even achieve a bit rate lowering the side information than the characterization 162. As shown in the figure, time grouping will form a high-multiple tuple (1, 0), allowing construction. Specify the shortest code character to the Huffman codebook of the previous tuple. As can be seen clearly in Figure 4, the high degree of resilience of the inventions using adaptive clustering and differential encoding allows for the selection of strategies that are best suited for the original audio, thus allowing the low side information bit rate to be maintained. In summary, in a preferred embodiment, the quantization parameter 値 is first differentially encoded for time (variable i) and frequency (variable 2). The parameters of the lock formation are then adapted to time (variable a) and frequency (variable b) to be grouped. As a result, there are four combinations (la' lb' 2a' 2b)' from which the best is selected and the decoder is signaled. This may only be represented by, for example, a bit combination 〇〇, 〇 1,1 〇,! i The binary information of the variables la, lb, 2a, 2b is completed. Figure 5 is a diagram showing decoding of the encoded parameter region by a decoder according to the present invention, wherein the parameter region includes a first -17-1321316 frame set having a first spectral parameter set and a second spectral parameter set Frame. The decoder 200 includes a decompressor 202 and a frame builder 204. The decompressor receives the encoded parameter region 206 on the input to obtain a parameter tuple sequence 208. The parameter tuple sequence 208 is input to the frame builder 204. The frame builder appends a receive sequence indicator 210 indicating what sequence of tuples the encoder has used to construct the encoded parameter region. The frame builder 204 then rearranges the tuple sequence 208 directed by the sequence indicator 210, reconstructing the first frame 112a and the second frame 112b from the parameter tuple sequence 208. The above-described preferred embodiment of the present invention achieves further enhancement of coding performance by using a multi-dimensional Huffman code to introduce an adaptive segmentation of the code to be encoded. As an example, two-dimensional groupings for frequency and two-dimensional groupings for time can be completed. The encoding structure then completes the encoding of the two types and selects a more advantageous one (i.e., a variant requiring fewer bits). The decision signal is sent to the decoder through the side information. As illustrated in Figure 3, in another example, it is also possible to construct a higher dimensional Huffman code and apply a different grouping strategy to construct the tuple. This hypothetical example indicates that only the parameters from two consecutive frames are grouped together to construct a grouping strategy for the tuple. It is also possible to use groups of three or more consecutive frames to group them in a direct manner. In the revision of the inventive encoder, it is also possible to use different Huffman codebooks, combined differential grouping and differential coding strategies to obtain the shortest possible side information representation. This may be at the expense of additional side information parameters, which in turn reduces the information bit rate at the side of the encoded audio, and signals the Huffman codebook -18-1321316 used for encoding. The preferred embodiment illustrated by the present invention represents, for example, an inventive concept in which the grouping policy is unchanged within two consecutive time frames. In the revision of the present invention, it is of course possible to have multiple changes between time and frequency grouping within the two frames, which means that a sequence indicator is also provided within the frame, and a signal of the group policy change is issued. In the hypothetical example, the parameter is differentially encoded before Huffman coding. Of course, before the Huffman coding of the parameters, every other lossless coding rule is also possible. The goal of the coding is to obtain the same set of tuples as possible. There are four different parameter characterizations in Figure 4, namely, absolute characterization, frequency differential characterization, time differential characterization, and time and frequency differential characterization. Select among the four characterizations, as shown in the fourth figure, the side information that signals the characterization used must be at least two bits. In order to balance the possible coding performance benefits for additional spectral characterization, it is of course possible to decide in principle that only two possible characterizations are allowed, and the reduced spectral representation is expressed as a single bit length. As an example of an inventive decoder, Fig. 5 shows a decoder 200 that additionally receives certain side information by dividing the coding parameter area 206. In the hypothetical example, the 'guide frame builder 204 can of course handle any other side information needed', particularly a spectral representation representation, indicating that the spectral representation of the original frame is used. The method of the invention may be carried out in a hard or soft manner, depending on certain implementation requirements of the method of the invention. The implementation can use a digitally stored disk with an electronically readable control signal, and a DVD or CD' ride cooperates with a programmable computer system to enable the method of the present invention to be implemented at -19-1321316. Usually, when the machine can read the computer program product of the vehicle, the code acts on the method of the present invention and is therefore implemented for the computer program to execute the computer program. Although the specific embodiment of the parameter has been made, those skilled in the art will Other changes are made without departing from the spirit and the embodiment of the invention, without departing from the general inventive concept. [Simple diagram of the drawing] The first graph is an inventive compression by the parameter drawing, and the third graph shows two possibilities. The third graph is an additional possibility. The fourth graph is a differential encoding structure. The fifth graph is The sixth diagram of the inventive decoder is the first -v / - 刖 technology [representative of the main part of the long f number description] 60 joint - _jL - body sound device 90 compression unit 100 feeder The present invention is therefore a code storage product When the computer is executed on a computer, the method of the present invention is implemented. In other words, the computer program has a program code that is less than one of the invention methods on the computer. The above items are specifically indicated and described to be various in form and detail. It is to be understood that a preferred embodiment of the invention encompassed by the different applications and the following application items can be found, wherein: element: suitable grouping parameter; adaptive grouping parameter; channel encoder 1321316

102 位元估算器 104 供應器 10 6a 元組 106b 元組 108a 資料輸入 1 08b 資料輸入 1 10a 資料輸入 1 10b 資料輸入 112a 發訊輸出 118 編碼區 120a 輸出 120b 發訊線 124a 編碼區 1 24b 編碼區 1 30a - 130d 時框 134a 元組 134b 元組 13 6a 元組 138 元組 140 元組 142 碼字元 144 碼字元 14 6a 二維霍夫曼碼 146b 三維霍夫曼碼 -21 1321316 14 6c 148a 148b 150b 152a 152b 160 162102 bit estimator 104 provider 10 6a tuple 106b tuple 108a data input 1 08b data input 1 10a data input 1 10b data input 112a signaling output 118 coding area 120a output 120b transmission line 124a coding area 1 24b coding area 1 30a - 130d block 134a tuple 134b tuple 13 6a tuple 138 tuple 140 tuple 142 code character 144 code character 14 6a 2D Huffman code 146b 3D Huffman code-21 1321316 14 6c 148a 148b 150b 152a 152b 160 162

164 200 202 204 206 208 2 10164 200 202 204 206 208 2 10

112a 1 12b 四維霍夫曼碼 頻Φ 時間 時間元組 時間分群元組 時間分群元組 參數 差分編碼參數 表徵 表徵 解碼器 解壓縮器 訊框架構器 參數區 參數元組序列 序列標式 第一訊框 第二訊框 -22-112a 1 12b four-dimensional Huffman code frequency Φ time time tuple time grouping tuple time grouping tuple parameter differential coding parameter characterization characterization decoder decompressor frame structure parameter parameter parameter tuple sequence sequence standard first frame Second frame-22-

1321316 * 魯丨4夕曰修正替換頁 ' 第95112773號「參數壓縮用之壓縮單元及方法、解碼已編 碼參數區之解碼器及方法、以及電腦程式」專利案 ( 2009年12月7日修正） 十、申請專利範圍： 1· 一種參數壓縮用之壓縮單元，該參數包含代表原始信號 之第一部位的第一參數集，該參數更包含代表該原始信 號之第二部位的第二參數集，該第二部位毗鄰該第一部 位，其包含： ® 用於供給第一元組（TUPLE )和第二元組之供給器，各元 組具有至少兩個參數，該第一元組之兩個參數來自第一 參數集而第二元組之一個參數來自第一參數集且一個參 數來自第二參數集； 位元估算器，係依據編碼規則用以估算使用含第一元 組之第一元組序列將參數及加以編碼，以及使用含第二 元組之第二元組序列將參數及加以編碼所需之位元數； 以及 用於供應已編碼區之供應器，該供應器可作用爲使用 形成較低位元數之元組序列，供應已編碼區，並用於供 應表示元組序列之序列標示，其中之編碼區係得自該元 組序列。 2.如申請專利範圍第1項之壓縮單元，其中該第_•和第二 參數集爲包含原始信號之第一和第二部位所代表之頻譜 3.如申請專利範圍第1項之壓縮單元，其中在時間或空間 1321316 _ ' 曰修正替換頁 ' 上，該原始信號之第一部位和第二部位毗鄰。 4. 如申請專利範圍第1項之壓縮單元，其中該第一參數集 及包含原始音訊或視訊信號之第一訊框的表徵，且其中 該等二參數集包含原始音訊或視訊信號之第二訊框的表 徵。 5. 如申請專利範圍第1項之壓縮單元，其中該供給器可作 用爲只使用其參數是來自一個單一參數集元組，藉以供 給第一元組及第一序列元組中所有剩下之元組，並只使 ^ 用其至少一個參數是來自第一參數集而至少一個參數是 來自第二參數集之元組，藉以供給第一元組及第二序列 元組中所有剩下之元組：以及 其中該供應器可作用爲供應一己編碼區之一單一序列 標不。 6. 如申請專利範圍第1項之壓縮單元，其中該供給器可作 用爲供給由來自一個單一參數集之兩個參數組成的第一 元組，該兩個參數在原始信號之表徵之中爲毗鄰參數； I 以及供給由來自第一參數集之第一參數和來自第二參數 集之第二參數組成的第二元組，該第一和第二參數在該 表徵之中說明該原始信號之相同特性。 7. 如申請專利範圍第1項之壓縮單元，其中該供給器可作 用爲供給由來自一個單一參數集之兩個參數的整數倍組 成的第一元組，該參數在原始信號之中爲毗鄰參數；以 及 供給由來自第一參數集之兩個或更多個參數組成的第二 1321316 納夕日修正替换頁 一 _ ~~ -. 元組’該參數在原始信號和來自第二參數組之相同參數 個數的表徵之中爲毗鄰，該參數在原始信號之表徵之中 爲毗鄰’來自第一和第二參數集之參數在該表徵之中說 明原始信號之相同特性。 8. 如申請專利範圍第1項之壓縮單元，其中該供給器可作 用爲供給第一元組，該第一元組由代表來自一參數集之 原始信號的奇數個毗鄰參數組成；以及 供給第二元組，該第二元組由代表來自一參數集之原始 信號的多數個毗鄰參數及由代表來自另一參數集之原始 信號的少數個毗鄰參數組成，其中爲了使第二序列元組 爲連續元組，該多數個參數另外取自第一和第二參數集 〇 9. 如申請專利範圍第1項之壓縮單元，其中該位元估算器 可作用爲使用編碼規則，將第一序列元組加以編碼，得 到第一編碼區並將第二序列元組加以編碼，得到第二編 碼區，並藉由計算第一和第二編碼區之位元來估算必要 之位元數；以及 其中該供應器可作用爲推進第一或第二編碼區，選取較 低位元數區。 10. 如申請專利範圍第1項之壓縮單元， 其中該位元估算器可作用爲依據編碼規則，估算將元組 序列編碼所需之位元；以及 其中該供應器可作用爲供應較低位元數給已編碼區，將 元組序列加以編碼。 1321316 夕日修正替換頁 11. 如申請專利範圍第1項之壓縮單元，其中該參數集包含 代表視訊信號或音訊信號之參數。 12. 如申請專利範圍第1項之壓縮單元，其中該編碼規則爲 具有相同長度但不同參數之編碼元組係能形成不同長度 之編碼字元。 13. 如申請專利範圍第1項之壓縮單元，其中該參數包含說 明第一和第二原始音訊聲道之間之空間相互間關係的雙 聲道佇列編碼（BCC )參數且其中該BCC參數是選自以下 之BCC參數清單： 相互間聲道之一致性/相關性（I CC); 相互間聲道之位準差（ICLD); 相互間聲道之時間差（ICTD); 相互間聲道之相位差（IPD)。 14. 如申請專利範圍第1項之壓縮單元，更包含差分編碼器 ，用以處理原始信號，使得第一參數集和第二參數集包 含差異表徵，該差異爲時間差、頻率差或時間及頻率差 〇 15. 如申請專利範圍第14項之壓縮單元， 其中該供給器之元組可作用在供給第一元組族群和第二 元組族群，該元組族群之中之元組的參數是來自下列表 徵之一： 時間差表徵、頻率差表徵、時間和頻率差表徵及絕對 表徵， 其中該第一元組族群的兩個參數來自第—參數集且其 -4 - ⑶ 1316 伽碩夕曰修正替換頁 中該第二元組族群的一個參數來自第一參數集而一個參 數來自第二參數集；以及 其中該位元估算器可作用在估算將至少八個元組序列 之元組加以編碼所需之位元數’各序列爲元組族群之其 中一個元組；以及 其中，除序列標示外，該供給器可作出在發出信號給 表徵標示，該表徵標示表示形成較低位元數之表徵。 W·如申請專利範圍第15項之壓縮單元，其中該供給器可作 用爲使用具有至少三個位元之二進位資料字元來發出序 列標示和表徵標示的信號。 如申請專利範圍第1項之壓縮單元，其中該編碼規則爲 各元組具有隨之相關之不同編碼字元族群的單一編碼字 元。 18.如申請專利範圍第1項之壓縮單元，其中該編碼規則包 含霍夫曼編碼簿（Huffman codebook)。 19_ —種用於將已編碼之參數區加以解碼之解碼器，該參數 包含代表原始信號之第一部位的第一參數集，該參數更 包含代表原始信號之第二部位的第二參數集，該第二部 位毗鄰該第一部位，並用於處理序列標示，其包含 解壓縮器，該解壓縮器可作用爲使用依據用以將元組 序列編碼之編碼規則，將已編碼之參數區加以解壓縮而 得到參數之元組序列，各該元組具有至少兩個參數；以 及 用於接收序列標示之訊框建構器，該序列標示表示來 1321316 钟Ν'上月夕日修正替換頁 自已編碼區下方之多數不同序列的已用過元組序列，並 使用已用過之元組序列的資訊加以建構參數集。 20·如申請專利範圍第19項之解碼器，其中該第一和第二參 數集包含原始信號之第一和第二部位所代表之頻譜。 2 1 ·如申請專利範圍第1 9項之解碼器，其中在時間或空間上 ，該原始信號之第一部位和第二部位毗鄰。 22. 如申請專利範圍第19項之解碼器，其中該第一參數集包 含原始音訊或視訊信號之第一訊框的表徵且其中該第二 參數集包含原始音訊或視訊信號之第二訊框的表徵。 23. 如申請專利範圍第19項之解碼器，其中該訊框建構器可 作用爲一個已編碼之參數區接收一單一序列標示，並藉 由使用只包含由序列標示所表示形式之元組的元組序列 加以建構參數集。 24. 如申請專利範圍第19項之解碼器，其中該訊框建構器可 作用爲建構包含代表視訊信號或音訊信號之參數的參數 集。 25. 如申請專利範圍第19項之解碼器，其中該訊框建構器可 作用爲建構參數集，該參數集包含說明第一和第二原始 音訊聲道之間之空間相互間關係的雙聲道佇列編碼（BCC) 參數且其中該BCC參數是選自以下之BCC參數清單： 相互間聲道之一致性/相關性（丨CC); 相互間聲道之位準差（I CLD); 相互間聲道之時間差（I CTD ); 相互間聲道之相位差（I PD)。 1321316 —»^1 · t 彿/上月j?日修正替換頁 26. 如申請專利範圍第19項之解碼器，更包含用於接收表徵 標示並用於處理參數集使得從差異性之頻譜表徵得到該 參數之差分解碼器，該差異性爲時間差、頻率差或時間 和頻率差，其中該差分解碼器可作用爲依據該表徵標示 加以處理第一和第二參數集。 27. 如申請專利範圍第26項之解碼器，其中該訊框建構器及 差分解碼器可作用爲接收序列標示和表徵標示來作爲具 有至少三個位元之二進位資料字元^ 28. 如申請專利範圍之第19項之解碼器》其中該解碼規則爲 在已編碼參數區之中的各編碼字元具有隨之相關不同元 組族群的單一元組。 29. —種參數壓縮用之方法’該參數包含代表原始信號之第 一參數集，該參數更包含代表原始信號之第二部位的第 二參數集，該第二部位毗鄰該第一部位，該方法包含： 供給該第一元組和該第二元組，各該元組具有至少兩 個參數，該第一元組之兩個參數來自第一參數集，而該 第二元組之一個參數來自第一參數集且一個參數來自第 二參數集； 依據編碼規則估算使用含第一元組之第一元組序列將 參數及加以編碼’以及使用含第二元組之第二元組序列 .將訊框加以編碼； 使用形成較低位元數之元組序列，供應已編碼區；以 及 供應表示元組序列之序列標示，其中之編碼區係得自 1321316 _ • 辦7印夕日修正替換頁 該元組序列。 30·—種已編碼之參數區解碼用之方法，該參數包含代表原 始信號之第一部位的第一參數集，該參數更包含代表原 始信號之第二部位的第二參數集，該第二部位毗鄰該第 一部位，並用於處理序列標示，該方法包含： 使用依據將元組序列編碼所用之編碼規則的解碼規則 ，將已編碼之參數區加以解壓縮而得到參數之元組序列 ，各該元組具有至少兩個參數； ® 接收序列標示，該序列標示表示來自已編碼區下方之 許多不同序列的已用過元組序列；以及 使用已用過之元組序列的資訊加以建構參數集。 31.—種具有當在電腦上執行時用以實施壓縮參數用之方法 的程式碼之電腦程式，該參數包含代表原始信號之第一 部位的第一參數集，該參數更包含代表原始信號之第二 部位的第二參數集，該第二部位毗鄰該第一部位，該方 法包含： ^ 供給第一元組和第二元組，各元組具有至少兩個參數 ，第一元組之兩個參數來自第一參數集而第二元組之一 個參數來自第一參數集且一個參數來自第二參數集； 依據編碼規則用以估算使用含第一元組之第一元組序 列將參數集及加以編碼，以及使用含第二元組之第二元 組序列將訊框加以編碼所需之位元數； 使用形成較低位元數之元組序列供應已編碼區；以及 供應表示元組序列之序列標示，其中之編碼區係得自 1321316 會日修正替換頁 該元組序列。 32.—種具有當在電腦上執行時用以實施已編碼參數已解碼 用之方法的程式碼之電腦程式，該參數包含第一參數集 ，該參數集包含代表原始信號之第一部位的第一參數集 ，該參數更包含代表原始信號之第二部位的第二參數集 ，該第二部位毗鄰第一部位，並用以處理序列標示，該 方法包含； 依據將元組序列編碼所用之編碼規則的解碼規則，將 已編碼之參數區加以解壓縮所而得到參數之元組序列， 各該元組具有至少兩個參數； 接收序列標示，該序向標示表示來自已編碼區下方之 許多不同序列的已用過元組序列；以及 使用已用過之元組序列的資訊加以建構參數集。1321316 * Reckless 4th Anniversary Correction Replacement Page 'No. 95112773 "Compression Unit and Method for Parameter Compression, Decoder and Method for Decoding Coded Parameter Area, and Computer Program" Patent Case (Revised on December 7, 2009) X. Patent application scope: 1. A compression unit for parameter compression, the parameter comprising a first parameter set representing a first portion of the original signal, the parameter further comprising a second parameter set representing a second portion of the original signal, The second portion is adjacent to the first portion and includes: ® a feeder for supplying a first tuple (TUPLE) and a second tuple, each tuple having at least two parameters, two of the first tuple The parameter is from the first parameter set and one parameter of the second tuple is from the first parameter set and one parameter is from the second parameter set; the bit estimator is used to estimate the first element using the first tuple according to the encoding rule The group sequence encodes and encodes the parameters, and uses the second tuple sequence containing the second tuple to encode and encode the parameters; and the provider for supplying the encoded region, Supply action may be formed using a sequence of tuples of a lower number of bits, the coding region has been supplied, and to supply a sequence of tuples marked, wherein the coding region were obtained from the sequence of tuples. 2. The compression unit of claim 1, wherein the first and second parameter sets are spectra comprising the first and second portions of the original signal. 3. The compression unit of claim 1 , wherein the first portion and the second portion of the original signal are adjacent to each other on the time or space 1321316 _ ' 曰 correction replacement page ′. 4. The compression unit of claim 1, wherein the first parameter set and the first frame containing the original audio or video signal, and wherein the two parameter sets comprise the original audio or the second video signal Characterization of the frame. 5. The compression unit of claim 1, wherein the feeder is operable to use only a parameter from a single parameter set tuple, thereby supplying all remaining in the first tuple and the first sequence tuple a tuple, and only uses at least one parameter from the first parameter set and at least one parameter is a tuple from the second parameter set, thereby supplying all remaining elements in the first tuple and the second sequence tuple Group: and wherein the provider can act as a single sequence indicator for one of the encoded regions. 6. The compression unit of claim 1, wherein the feeder is operable to supply a first tuple consisting of two parameters from a single parameter set, the two parameters being among the representations of the original signal Adjacent parameters; I and a second tuple consisting of a first parameter from the first parameter set and a second parameter from the second parameter set, the first and second parameters indicating the original signal in the representation The same characteristics. 7. The compression unit of claim 1, wherein the feeder is operable to supply a first tuple consisting of integer multiples from two parameters of a single parameter set, the parameter being adjacent to the original signal a parameter; and a second 1321316, which is composed of two or more parameters from the first parameter set, replaces the page _~~ -. tuple 'this parameter is the same as the original signal and the second parameter group Among the characterizations of the number of parameters is the contiguousness, which parameter is adjacent to the characterization of the original signal. The parameters from the first and second parameter sets illustrate the same characteristics of the original signal in the characterization. 8. The compression unit of claim 1, wherein the feeder is operable to supply a first tuple consisting of an odd number of adjacent parameters representing an original signal from a parameter set; and a supply a two-tuple consisting of a plurality of contiguous parameters representing the original signal from a parameter set and a few contiguous parameters representing the original signal from another parameter set, wherein in order to make the second sequence tuple a contiguous tuple, the majority of the parameters are additionally taken from the first and second parameter sets 〇 9. As in the compression unit of claim 1, wherein the bit estimator can function to use the encoding rule, the first sequence element Encoding the group to obtain a first coding region and encoding the second sequence tuple to obtain a second coding region, and estimating a necessary number of bits by calculating bits of the first and second coding regions; and wherein the The supplier can act to advance the first or second coding region and select the lower bit number region. 10. The compression unit of claim 1, wherein the bit estimator is operable to estimate a bit required to encode the tuple sequence according to an encoding rule; and wherein the supplier is operable to supply a lower bit The number of elements is given to the coded area, and the sequence of the tuple is encoded. 1321316 夕日修正 replacement page 11. The compression unit of claim 1, wherein the parameter set contains parameters representative of a video signal or an audio signal. 12. The compression unit of claim 1, wherein the encoding rule is an encoding tuple having the same length but different parameters capable of forming encoding characters of different lengths. 13. The compression unit of claim 1, wherein the parameter comprises a two-channel constrained coding (BCC) parameter indicating a spatial relationship between the first and second original audio channels and wherein the BCC parameter It is a list of BCC parameters selected from the following: Consistency/correlation of channels between each other (I CC); Interval difference (ICLD) between channels; Inter-channel time difference (ICTD); Inter-channel Phase difference (IPD). 14. The compression unit of claim 1 further comprising a differential encoder for processing the original signal such that the first parameter set and the second parameter set comprise a difference characterization, the difference being a time difference, a frequency difference, or a time and frequency Clause 15. The compression unit of claim 14, wherein the tuple of the feeder is operable to supply the first tuple group and the second tuple group, and the tuple parameter of the tuple group is From one of the following characterizations: time difference characterization, frequency difference characterization, time and frequency difference characterization, and absolute characterization, where the two parameters of the first tuple group are from the first parameter set and its -4 - (3) 1316 gamma correction One parameter of the second tuple group in the replacement page is from the first parameter set and one parameter is from the second parameter set; and wherein the bit estimator is operable to estimate the tuple encoding at least eight tuple sequences The number of bits required 'each sequence is one of the tuples of the tuple group; and wherein, in addition to the sequence indication, the supplier can make a signal to indicate the signature, This characterization mark indicates a representation that the number of lower bits is formed. W. The compression unit of claim 15 wherein the feeder is operative to use a binary data character having at least three bits to issue a sequence indication and a characterization signature. The compression unit of claim 1, wherein the encoding rule is a single coded character of each tuple having a different code character group associated therewith. 18. The compression unit of claim 1, wherein the encoding rule comprises a Huffman codebook. a decoder for decoding an encoded parameter region, the parameter comprising a first parameter set representing a first portion of the original signal, the parameter further comprising a second parameter set representing a second portion of the original signal, The second portion is adjacent to the first portion and is for processing a sequence indicator, comprising a decompressor, the decompressor operable to use the encoding parameter according to the encoding sequence for encoding the tuple sequence to solve the encoded parameter region Compressing to obtain a sequence of tuples of parameters, each of the tuples having at least two parameters; and a frame constructor for receiving the sequence designation, the sequence mark indicating that the 1321316 clock Ν 'last moon day correction replacement page is below the coded area Most of the different sequences of used tuple sequences have been constructed using the information of the used tuple sequences. 20. The decoder of claim 19, wherein the first and second sets of parameters comprise a spectrum represented by the first and second portions of the original signal. 2 1 The decoder of claim 19, wherein the first portion and the second portion of the original signal are adjacent in time or space. 22. The decoder of claim 19, wherein the first parameter set comprises a representation of a first frame of the original audio or video signal and wherein the second parameter set comprises a second frame of the original audio or video signal Characterization. 23. The decoder of claim 19, wherein the frame constructor acts as an encoded parameter region to receive a single sequence designation and by using a tuple containing only the form represented by the sequence designation The tuple sequence is constructed to construct a parameter set. 24. The decoder of claim 19, wherein the frame builder is operable to construct a parameter set comprising parameters representative of a video signal or an audio signal. 25. The decoder of claim 19, wherein the frame builder is operable to construct a parameter set, the parameter set comprising a double sound indicating a spatial relationship between the first and second original audio channels a track-column coding (BCC) parameter and wherein the BCC parameter is a list of BCC parameters selected from the group consisting of: consistency/correlation of channels between each other (丨CC); level difference between the channels (I CLD); The time difference between the channels of each other (I CTD ); the phase difference (I PD) between the channels. 1321316 —»^1 · t Buddha / last month j? day correction replacement page 26. The decoder of claim 19 of the patent scope, further includes for receiving the characterization mark and for processing the parameter set so as to obtain the spectral characterization from the difference The differential decoder of the parameter, the difference being a time difference, a frequency difference, or a time and frequency difference, wherein the differential decoder is operative to process the first and second parameter sets in accordance with the characterization signature. 27. The decoder of claim 26, wherein the frame constructor and the differential decoder are operable to receive the sequence indicator and the characterization indicator as a binary data character having at least three bits. The decoder of claim 19, wherein the decoding rule is that each of the coded characters in the encoded parameter region has a single tuple associated with the different tuple group. 29. A method for parameter compression 'This parameter includes a first parameter set representing an original signal, the parameter further comprising a second parameter set representing a second portion of the original signal, the second portion being adjacent to the first portion, The method includes: supplying the first tuple and the second tuple, each tuple having at least two parameters, two parameters of the first tuple are from a first parameter set, and one parameter of the second tuple From the first parameter set and one parameter from the second parameter set; estimating and encoding the parameter using the first tuple sequence containing the first tuple according to the encoding rule' and using the second tuple sequence containing the second tuple. Encoding the frame; supplying the encoded region using a tuple sequence forming a lower number of bits; and supplying a sequence indication indicating the sequence of the tuple, wherein the coding region is obtained from 1321316 _ • 7th day correction replacement page The tuple sequence. 30. A method for decoding a coded parameter region, the parameter comprising a first parameter set representing a first portion of the original signal, the parameter further comprising a second parameter set representing a second portion of the original signal, the second The portion is adjacent to the first portion and is used to process the sequence indication. The method includes: decompressing the encoded parameter region using a decoding rule according to an encoding rule used to encode the tuple sequence to obtain a tuple sequence of parameters, each The tuple has at least two parameters; ® receives a sequence designation indicating the used tuple sequence from a number of different sequences below the coded region; and constructs a parameter set using information from the used tuple sequence . 31. A computer program having a code for implementing a method for compressing parameters when executed on a computer, the parameter comprising a first parameter set representing a first portion of the original signal, the parameter further comprising a representative of the original signal a second parameter set of the second portion, the second portion is adjacent to the first portion, the method comprising: ^ supplying a first tuple and a second tuple, each tuple having at least two parameters, two of the first tuple The parameters are from the first parameter set and one parameter of the second tuple is from the first parameter set and one parameter is from the second parameter set; the parameter set is used to estimate the parameter set using the first tuple sequence containing the first tuple And encoding, and using the number of bits required to encode the frame using a second tuple sequence of the second tuple; supplying the encoded region using a tuple sequence forming a lower number of bits; and supplying the representation tuple The sequence of the sequence is indicated, wherein the coding region is obtained from the 1321316 meeting date correction replacement page. 32. A computer program having a code for implementing a method for decoding encoded parameters when executed on a computer, the parameter comprising a first parameter set including a first portion representing a first portion of the original signal a parameter set, the parameter further comprising a second parameter set representing a second portion of the original signal, the second portion being adjacent to the first portion and for processing the sequence indication, the method comprising: encoding rules according to encoding the tuple sequence Decoding rules, decompressing the encoded parameter regions to obtain a tuple sequence of parameters, each tuple having at least two parameters; receiving a sequence indication indicating a plurality of different sequences from below the coded region The used tuple sequence; and the information set using the used tuple sequence to construct the parameter set.

• 2005
  • 2005-10-05 US US11/243,915 patent/US7991610B2/en active Active
• 2006
  • 2006-02-13 DK DK19156694.2T patent/DK3503409T3/en active
  • 2006-02-13 KR KR1020077022723A patent/KR100954181B1/en active Active
  • 2006-02-13 PT PT06723036T patent/PT1869774T/en unknown
  • 2006-02-13 BR BR122019014628-4A patent/BR122019014628B1/en active IP Right Grant
  • 2006-02-13 PL PL19156694.2T patent/PL3503409T3/en unknown
  • 2006-02-13 MX MX2007012663A patent/MX2007012663A/en active IP Right Grant
  • 2006-02-13 AU AU2006233512A patent/AU2006233512C1/en active Active
  • 2006-02-13 HU HUE19156694A patent/HUE069108T2/en unknown
  • 2006-02-13 PL PL06723036T patent/PL1869774T3/en unknown
  • 2006-02-13 BR BRPI0612356-2A patent/BRPI0612356B1/en active IP Right Grant
  • 2006-02-13 EP EP19156694.2A patent/EP3503409B1/en active Active
  • 2006-02-13 RU RU2007141934/09A patent/RU2368074C2/en active
  • 2006-02-13 BR BR122019014976-3A patent/BR122019014976B1/en not_active IP Right Cessation
  • 2006-02-13 FI FIEP19156694.2T patent/FI3503409T3/en active
  • 2006-02-13 PT PT191566942T patent/PT3503409T/en unknown
  • 2006-02-13 JP JP2008505744A patent/JP5027799B2/en active Active
  • 2006-02-13 EP EP06723036.7A patent/EP1869774B1/en active Active
  • 2006-02-13 CN CN2006800120933A patent/CN101160726B/en active Active
  • 2006-02-13 ES ES19156694T patent/ES2993481T3/en active Active
  • 2006-02-13 CA CA2607460A patent/CA2607460C/en active Active
  • 2006-02-13 ES ES06723036T patent/ES2739531T3/en active Active
  • 2006-02-13 WO PCT/EP2006/001295 patent/WO2006108464A1/en active Application Filing
  • 2006-02-15 MY MYPI20060645A patent/MY141958A/en unknown
  • 2006-04-11 TW TW095112773A patent/TWI321316B/en active
• 2007
  • 2007-09-25 IL IL186315A patent/IL186315A/en active IP Right Grant
  • 2007-11-12 NO NO20075796A patent/NO343207B1/en unknown
• 2010
  • 2010-11-17 US US12/948,417 patent/US9043200B2/en active Active

RetroSearch is an open source project built by @garambo | Open a GitHub Issue

Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo

HTML: 3.2 | Encoding: UTF-8 | Version: 0.7.4